From: Harish Jenny K N <harish_kandiga@mentor.com>
To: Geert Uytterhoeven <geert@linux-m68k.org>, Rob Herring <robh@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>,
Peter Maydell <peter.maydell@linaro.org>,
QEMU Developers <qemu-devel@nongnu.org>,
Phil Reid <preid@electromag.com.au>,
Geert Uytterhoeven <geert+renesas@glider.be>,
Jonathan Corbet <corbet@lwn.net>,
Marc Zyngier <marc.zyngier@arm.com>,
Linus Walleij <linus.walleij@linaro.org>,
"open list:DOCUMENTATION" <linux-doc@vger.kernel.org>,
Magnus Damm <magnus.damm@gmail.com>,
Christoffer Dall <christoffer.dall@arm.com>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
Linux-Renesas <linux-renesas-soc@vger.kernel.org>,
Bartosz Golaszewski <bgolaszewski@baylibre.com>,
"open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS"
<devicetree@vger.kernel.org>, Alexander Graf <graf@amazon.com>,
"open list:GPIO SUBSYSTEM" <linux-gpio@vger.kernel.org>,
Paolo Bonzini <pbonzini@redhat.com>,
Eugeniu Rosca <erosca@de.adit-jv.com>
Subject: Re: [PATCH v3 4/7] dt-bindings: gpio: Add gpio-repeater bindings
Date: Thu, 16 Jan 2020 10:39:50 +0530 [thread overview]
Message-ID: <4dccaf48-c652-a5f4-49f3-ccc279d5548e@mentor.com> (raw)
In-Reply-To: <bb5fb539-0d0d-6356-35c2-8ba47cb9fcbf@mentor.com>
Hi Linus,
On 07/01/20 2:52 PM, Harish Jenny K N wrote:
> On 06/01/20 1:42 PM, Geert Uytterhoeven wrote:
>> Hi Rob,
>>
>> On Fri, Dec 6, 2019 at 4:04 PM Rob Herring <robh@kernel.org> wrote:
>>> On Fri, Dec 6, 2019 at 3:17 AM Geert Uytterhoeven <geert@linux-m68k.org> wrote:
>>>> On Thu, Dec 5, 2019 at 10:06 PM Rob Herring <robh@kernel.org> wrote:
>>>>> On Wed, Nov 27, 2019 at 09:42:50AM +0100, Geert Uytterhoeven wrote:
>>>>>> Add Device Tree bindings for a GPIO repeater, with optional translation
>>>>>> of physical signal properties. This is useful for describing explicitly
>>>>>> the presence of e.g. an inverter on a GPIO line, and was inspired by the
>>>>>> non-YAML gpio-inverter bindings by Harish Jenny K N
>>>>>> <harish_kandiga@mentor.com>[1].
>>>>>>
>>>>>> Note that this is different from a GPIO Nexus Node[2], which cannot do
>>>>>> physical signal property translation.
>>>>> It can't? Why not? The point of the passthru mask is to not do
>>>>> translation of flags, but without it you are always doing translation of
>>>>> cells.
>>>> Thanks for pushing me deeper into nexuses!
>>>> You're right, you can map from one type to another.
>>>> However, you cannot handle the "double inversion" of an ACTIVE_LOW
>>>> signal with a physical inverter added:
>>>>
>>>> nexus: led-nexus {
>>>> #gpio-cells = <2>;
>>>> gpio-map = <0 0 &gpio2 19 GPIO_ACTIVE_LOW>, // inverted
>>>> <1 0 &gpio2 20 GPIO_ACTIVE_HIGH>, // noninverted
>>>> <2 0 &gpio2 21 GPIO_ACTIVE_LOW>; // inverted
>>>> gpio-map-mask = <3 0>;
>>>> // default gpio-map-pass-thru = <0 0>;
>>>> };
>>>>
>>>> leds {
>>>> compatible = "gpio-leds";
>>>> led6-inverted {
>>>> gpios = <&nexus 0 GPIO_ACTIVE_HIGH>;
>>>> };
>>>> led7-noninverted {
>>>> gpios = <&nexus 1 GPIO_ACTIVE_HIGH>;
>>>> };
>>>> led8-double-inverted { // FAILS: still inverted
>>>> gpios = <&nexus 2 GPIO_ACTIVE_LOW>;
>>>> };
>>>> };
>>>>
>>>> It "works" if the last entry in gpio-map is changed to GPIO_ACTIVE_HIGH.
>>>> Still, the consumer would see the final translated polarity, and not the
>>>> actual one it needs to program the consumer for.
>>> I'm not really following. Why isn't a double inversion just the same
>>> as no inversion?
>> Because the nexus can only mask and/or substitute bits.
>> It cannot do a XOR operation on the GPIO flags.
>>
>>>>>> While an inverter can be described implicitly by exchanging the
>>>>>> GPIO_ACTIVE_HIGH and GPIO_ACTIVE_LOW flags, this has its limitations.
>>>>>> Each GPIO line has only a single GPIO_ACTIVE_* flag, but applies to both
>>>>>> th provider and consumer sides:
>>>>>> 1. The GPIO provider (controller) looks at the flags to know the
>>>>>> polarity, so it can translate between logical (active/not active)
>>>>>> and physical (high/low) signal levels.
>>>>>> 2. While the signal polarity is usually fixed on the GPIO consumer
>>>>>> side (e.g. an LED is tied to either the supply voltage or GND),
>>>>>> it may be configurable on some devices, and both sides need to
>>>>>> agree. Hence the GPIO_ACTIVE_* flag as seen by the consumer must
>>>>>> match the actual polarity.
>>>>>> There exists a similar issue with interrupt flags, where both the
>>>>>> interrupt controller and the device generating the interrupt need
>>>>>> to agree, which breaks in the presence of a physical inverter not
>>>>>> described in DT (see e.g. [3]).
>>>>> Adding an inverted flag as I've suggested would also solve this issue.
>>>> As per your suggestion in "Re: [PATCH V4 2/2] gpio: inverter: document
>>>> the inverter bindings"?
>>>> https://lore.kernel.org/linux-devicetree/CAL_JsqLp___2O-naU+2PPQy0QmJX6+aN3hByz-OB9+qFvWgN9Q@mail.gmail.com/
>>>>
>>>> Oh, now I understand. I was misguided by Harish' interpretation
>>>> https://lore.kernel.org/linux-devicetree/dde73334-a26d-b53f-6b97-4101c1cdc185@mentor.com/
>>>> which assumed an "inverted" property, e.g.
>>>>
>>>> inverted = /bits/ 8 <0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0>;
>>>>
>>>> But you actually meant a new GPIO_INVERTED flag, to be ORed into the 2nd
>>>> cell of a GPIO specifier? I.e. add to include/dt-bindings/gpio/gpio.h"
>>>>
>>>> /* Bit 6 expresses the presence of a physical inverter */
>>>> #define GPIO_INVERTED 64
>>> Exactly.
>> OK, makes sense.
>
> The reason I went for "inverted" property is because, we can specify this for gpios at provider side.
>
> The usecase needed to define the polarity which did not have kernel space consumer driver.
>
>
> I am not sure how do we achieve this using GPIO_INVERTED flag. We need some sort of node/gpio-hog to specify these
>
> type of properties? Otherwise gpio-pin will be held by kernel or the module using the hog property and the user space application will not be able to access pin.
>
>
> or please let me know if I am missing something.
>
>
>>>> We need to be very careful in defining to which side the GPIO_ACTIVE_*
>>>> applies to (consumer?), and which side the GPIO_INVERTED flag (provider?).
>>>> Still, this doesn't help if e.g. a FET is used instead of a push-pull
>>>> inverter, as the former needs translation of other flags (which the
>>>> nexus can do, the caveats above still applies, though).
>>> Yes. Historically the cells values are meaningful to the provider and
>>> opaque to the consumer. Standardized cell values changes that
>>> somewhat. I think we want the active flag to be from the provider's
>>> prospective because the provider always needs to know. The consumer
>>> often doesn't need to know. That also means things work without the
>>> GPIO_INVERTED flag if the consumer doesn't care which is what we have
>>> today already and we can't go back in time.
>>>
> Things will work without GPIO_INVERTED flag for consumers which can specify GPIO_ACTIVE_* flags.
>
>
>
>>>> Same for adding IRQ_TYPE_INVERTED.
>>> I suppose so, yes.
>>>
>>>> Related issue: how to handle physical inverters on SPI chip select lines,
>>>> if the SPI slave can be configured for both polarities?
>>> Good question. Perhaps in a different way because we have to handle
>>> both h/w controlled and gpio chip selects.
>>>
>>> However, how would one configure the polarity in the device in the
>>> first place? You have to assert the CS first to give a command to
>>> reprogram it.
>> That's indeed true for a simple SPI slave.
>> But if it is a smarter device (e.g. a generic micro controller), it may use the
>> system's DTB to configure itself.
>>
>> Gr{oetje,eeting}s,
>>
>> Geert
>>
Can you please let me know your inputs on this ?
Now that Geert has sent v4 patch of GPIO Aggregator by "Dropping controversial GPIO repeater", I do not see the above mentioned inverter usecase can be handled anymore.
Is the observation/patch submitted in https://lore.kernel.org/linux-devicetree/dde73334-a26d-b53f-6b97-4101c1cdc185@mentor.com/ still not acceptable?
Thanks,
Harish
next prev parent reply other threads:[~2020-01-16 5:11 UTC|newest]
Thread overview: 48+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-11-27 8:42 [PATCH v3 0/7] gpio: Add GPIO Aggregator/Repeater Geert Uytterhoeven
2019-11-27 8:42 ` [PATCH v3 1/7] gpiolib: Add GPIOCHIP_NAME definition Geert Uytterhoeven
2019-11-28 3:38 ` Ulrich Hecht
2019-12-02 21:17 ` Eugeniu Rosca
2019-12-12 10:37 ` Linus Walleij
2019-11-27 8:42 ` [PATCH v3 2/7] gpiolib: Add support for gpiochipN-based table lookup Geert Uytterhoeven
2019-11-28 3:38 ` Ulrich Hecht
2019-12-12 13:20 ` Linus Walleij
2019-12-12 13:33 ` Geert Uytterhoeven
2019-12-12 14:36 ` Linus Walleij
2019-11-27 8:42 ` [PATCH v3 3/7] gpiolib: Add support for GPIO line " Geert Uytterhoeven
2019-11-28 3:39 ` Ulrich Hecht
2019-12-12 13:40 ` Linus Walleij
2019-11-27 8:42 ` [PATCH v3 4/7] dt-bindings: gpio: Add gpio-repeater bindings Geert Uytterhoeven
2019-11-28 3:39 ` Ulrich Hecht
2019-12-03 5:51 ` Harish Jenny K N
2019-12-05 21:06 ` Rob Herring
2019-12-06 9:17 ` Geert Uytterhoeven
2019-12-06 15:03 ` Rob Herring
2020-01-06 8:12 ` Geert Uytterhoeven
2020-01-07 9:22 ` Harish Jenny K N
2020-01-16 5:09 ` Harish Jenny K N [this message]
2019-11-27 8:42 ` [PATCH v3 5/7] gpio: Add GPIO Aggregator/Repeater driver Geert Uytterhoeven
2019-11-27 14:15 ` Eugeniu Rosca
2019-11-27 14:33 ` Geert Uytterhoeven
2019-11-28 3:40 ` Ulrich Hecht
2019-12-03 5:42 ` Harish Jenny K N
2019-12-03 8:17 ` Geert Uytterhoeven
2019-12-03 8:51 ` Harish Jenny K N
2019-12-03 10:51 ` Eugeniu Rosca
2020-01-09 13:35 ` Geert Uytterhoeven
2020-01-09 13:49 ` Eugeniu Rosca
2019-12-12 14:34 ` Linus Walleij
2019-12-12 15:24 ` Geert Uytterhoeven
2020-01-04 0:38 ` Linus Walleij
2020-01-06 8:23 ` Geert Uytterhoeven
2020-01-08 23:12 ` Linus Walleij
2019-11-27 8:42 ` [PATCH v3 6/7] docs: gpio: Add GPIO Aggregator/Repeater documentation Geert Uytterhoeven
2019-11-28 3:41 ` Ulrich Hecht
2019-12-12 14:42 ` Linus Walleij
2019-12-12 14:48 ` Geert Uytterhoeven
2020-01-04 0:21 ` Linus Walleij
2020-01-06 8:06 ` Geert Uytterhoeven
2019-11-27 8:42 ` [PATCH v3 7/7] MAINTAINERS: Add GPIO Aggregator/Repeater section Geert Uytterhoeven
2019-12-03 5:38 ` Harish Jenny K N
2020-01-18 1:46 ` [PATCH v3 0/7] gpio: Add GPIO Aggregator/Repeater Eugeniu Rosca
2020-01-20 9:33 ` Geert Uytterhoeven
2020-01-20 12:14 ` Eugeniu Rosca
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4dccaf48-c652-a5f4-49f3-ccc279d5548e@mentor.com \
--to=harish_kandiga@mentor.com \
--cc=bgolaszewski@baylibre.com \
--cc=christoffer.dall@arm.com \
--cc=corbet@lwn.net \
--cc=devicetree@vger.kernel.org \
--cc=erosca@de.adit-jv.com \
--cc=geert+renesas@glider.be \
--cc=geert@linux-m68k.org \
--cc=graf@amazon.com \
--cc=linus.walleij@linaro.org \
--cc=linux-doc@vger.kernel.org \
--cc=linux-gpio@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-renesas-soc@vger.kernel.org \
--cc=magnus.damm@gmail.com \
--cc=marc.zyngier@arm.com \
--cc=mark.rutland@arm.com \
--cc=pbonzini@redhat.com \
--cc=peter.maydell@linaro.org \
--cc=preid@electromag.com.au \
--cc=qemu-devel@nongnu.org \
--cc=robh@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).